This study aimed to build machine learning prediction models for predicting pathological subtypes of prevascular mediastinal tumors (PMTs). The candidate predictors were clinical variables and dynamic contrast–enhanced MRI (DCE-MRI)–derived perfusion parameters. The clinical data and preoperative DCE–MRI images of 62 PMT patients, including 17 patients with lymphoma, 31 with thymoma, and 14 with thymic carcinoma, were retrospectively analyzed. Six perfusion parameters were calculated as candidate predictors. Univariate receiver-operating-characteristic curve analysis was performed to evaluate the performance of the prediction models. A predictive model was built based on multi-class classification, which detected lymphoma, thymoma, and thymic carcinoma with sensitivity of 52.9%, 74.2%, and 92.8%, respectively. In addition, two predictive models were built based on binary classification for distinguishing Hodgkin from non-Hodgkin lymphoma and for distinguishing invasive from noninvasive thymoma, with sensitivity of 75% and 71.4%, respectively. In addition to two perfusion parameters (efflux rate constant from tissue extravascular extracellular space into the blood plasma, and extravascular extracellular space volume per unit volume of tissue), age and tumor volume were also essential parameters for predicting PMT subtypes. In conclusion, our machine learning–based predictive model, constructed with clinical data and perfusion parameters, may represent a useful tool for differential diagnosis of PMT subtypes.

Free-running cardiac quantitative susceptibility mapping allows for accurate detection and quantification of intramyocardial hemorrhage in the setting of acute myocardial infarction.

The developed delayed-phase dynamic contrast-enhanced MRI model for myocardial tissue assessment shows potential to enhance lesion contrast, quantify clinically relevant physiologic parameters, and support comprehensive evaluation of myocardial injury in heart disease.

Background/Objectives: Accurate prognostication for interstitial lung disease (ILD) remains challenging, limiting clinicians’ ability to optimize patient management strategies. This study aimed to evaluate the prognostic value of baseline quantitative CT-derived indices, including total lung volume (TLVcm³), normal lung volume% (NLV%), and fibrotic lung volume% (FLV%), for predicting three-year mortality in interstitial lung disease (ILD) patients. Methods: A total of 101 ILD patients were retrospectively enrolled. Baseline CT-derived indices, including TLVcm³, NLV% (−950 to −700 HU), and FLV% (−600 to +50 HU), were measured on chest CT. Baseline forced vital capacity(FVC)% predicted and diffuse capacity of lungs for carbon monoxide (DLCO)% predicted were collected. Survival analysis used Kaplan–Meier’s curves and log-rank tests. Uni- and multivariate Cox’s proportional hazards regression were performed. Pearson’s correlation was used between CT-derived indices, FVC% predicted, and DLCO% predicted. Results: During 3-year follow-up, 30 of 101 patients (29.70%) died. Deceased patients had a significantly lower baseline NLV% (59.27% ± 7.61% vs. 65.02% ± 7.82%, p = 0.001) and a higher FLV% (17.64% ± 7.98% vs. 13.34% ± 7.48%, p = 0.011) compared with survivors. Multivariate analysis identified baseline NLV% (adjusted hazard ratio 0.88, 95% CI: 0.78–0.99, p = 0.034) and DLCO% predicted (adjusted hazard ratio 0.97, 95% CI: 0.95–0.99, p = 0.007) as independent predictors of three-year mortality. Patients with NLV% ≤ 64.15 and FLV% ≥ 14.12 showed significantly worse survival outcomes (21.78% vs. 7.92%, p < 0.001;19.80% (20/101) vs. 9.90% (10/101), p < 0.001). CT-derived indices moderately correlated with FVC% predicted and DLCO% predicted. Conclusions: Baseline FLV% ≥ 14.12 and NLV% ≤ 64.15 can effectively stratify and differentiate outcomes in ILD patients. Baseline NLV% has the potential as a prognostic indicator for 3-year survival in ILD.